Interaction design in the pragmatic Web = reducing semiotic barriers to web-mediated communication

(1)

Heiko Horst Hornung

“Interaction Design in the Pragmatic Web –

Reducing Semiotic Barriers to Web-mediated

Collaboration.”

“Design da Interação na Web Pragmática –

Reduzindo Barreiras Semióticas na Colaboração

Mediada pela Web.”

CAMPINAS

2013

(2)

(3)

(4)

Ficha catalográfica

Universidade Estadual de Campinas

Biblioteca do Instituto de Matemática, Estatística e Computação Científica Maria Fabiana Bezerra Muller - CRB 8/6162

Hornung, Heiko Horst,

H785i HorInteraction design in the pragmatic Web - reducing semiotic barriers to Web-mediated communication / Heiko Horst Hornung. – Campinas, SP : [s.n.], 2013.

HorOrientador: Maria Cecília Calani Baranauskas.

HorTese (doutorado) – Universidade Estadual de Campinas, Instituto de Computação.

Hor1. Interação humano-computador. 2. Design de interação. 3. Semiótica e computação. I. Baranauskas, Maria Cecilia Calani,1954-. II. Universidade Estadual de Campinas. Instituto de Computação. III. Título.

Informações para Biblioteca Digital

Título em outro idioma: Design da interação na Web pragmática - reduzindo barreiras semióticas na colaboração mediada pela Web

Palavras-chave em inglês: Human-computer interaction Interaction design

Semiotics and computing

Área de concentração: Ciência da Computação Titulação: Doutor em Ciência da Computação Banca examinadora:

Maria Cecília Calani Baranauskas [Orientador] Cristiano Maciel

Rodrigo Bonacin

Ariadne Maria Brito Rizzoni Carvalho Maria Beatriz Felgar de Toledo Data de defesa: 11-12-2013

Programa de Pós-Graduação: Ciência da Computação

Powered by TCPDF (www.tcpdf.org)

(5)

(6)

(7)

Institute of Computing /Instituto de Computação University of Campinas /Universidade Estadual de Campinas

Interaction Design in the Pragmatic Web – Reducing

Semiotic Barriers to Web-mediated Collaboration.

Heiko Horst Hornung

1

December 11, 2013

Examiner Board/Banca Examinadora:

• Prof. Dr. Maria Cecília Calani Baranauskas (Supervisor/Orientadora) • Prof. Dr. Ariadne Maria Brito Rizzoni Carvalho

Institute of Computing - UNICAMP • Prof. Dr. Maria Beatriz Felgar de Toledo

Institute of Computing - UNICAMP • Prof. Dr. Cristiano Maciel

Institute of Computing - UFMT • Prof. Dr. Rodrigo Bonacin

Centro de Tecnologia da Informação Renato Archer (Substitute/Suplente) • Prof. Dr. Maria Teresa Eglér Mantoan

Faculty of Education - UNICAMP (Substitute/Suplente) • Dr. Maria Cecília Martins

Núcleo de Informática Aplicada a Educação - UNICAMP (Substitute/Suplente) • Prof. Dr. Sérgio Roberto Pereira da Silva

Department of Informatics - UEM

1_{Financial support: CAPES and CNPq scholarships}

(8)

(9)

Abstract

The Web and its underlying technologies enable interactions among people that were unimaginable a few years ago. An important type of purposeful interaction is collabora-tion. Mediated by the Web, people from different social and cultural backgrounds, with different needs, preferences and capabilities can collaborate with each other. Collabora-tion often takes place in heterogeneous contexts that are not only defined by the actual situations of the collaboration partners, but also by individual and collective past expe-riences. The Web as a medium has an impact on collaboration and facilitates or enables certain aspects of collaboration while making others more difficult.

In this PhD thesis we investigate Interaction Design related questions about web-mediated collaboration under a Pragmatic Web perspective. Our prime objective is to understand semiotic barriers to web-mediated collaboration and propose an approach to Interaction Design that reduces these barriers. Semiotic barriers are barriers related to communication, mediation and representation. These barriers emerge during web-based collaboration since many mechanisms of interpersonal face-to-face communication are not available. Depending on the context, semiotic barriers often have a negative impact on collaboration, but in some cases they might also have positive effects.

The approach to Interaction Design proposed in this PhD thesis is rooted in the Pragmatic Web and uses Organizational Semiotics and Activity Theory as its theoreti-cal and methodologitheoreti-cal frames of reference. The theoretic investigations were practitheoreti-cally grounded in real world practices by participating in a research project in the domain of inclusive education. We materialized the proposed approach in the design of a prototype and the implementation of the corresponding tool that supports a practice of inclusive education professionals. Furthermore we proposed and applied a pragmatics-driven eval-uation method in a longitudinal case study. Prototype design, tool implementation, and the conducted evaluation provided evidence that the proposed approach to pragmatics-driven Interaction Design can reduce semiotic barriers and thus promote web-mediated collaboration.

(10)

(11)

Resumo

A Web e suas tecnologias de base facilitam interações entre pessoas que alguns anos atrás não eram imagináveis. A colaboração é um tipo importante de interação que tem um propósito. Pessoas de diferentes contextos sociais e culturais, e com diferentes preferências e habilidades, podem colaborar mediadas pela Web. A colaboração muitas vezes acontece em contextos heterogêneos, que são definidos tanto pelas situações atuais dos parceiros na colaboração, quanto pelas experiências passadas, sejam elas individuais ou coletivas. A Web como um meio/uma mídia tem um impacto na colaboração e facilita certos aspectos da colaboração enquanto dificulta outros.

Adotando uma perspectiva informada pela Web Pragmática, nesta tese investigamos questões da colaboração mediada pela Web, relacionadas com o Design da Interação. Nosso objetivo principal é entender barreiras semióticas da colaboração mediada pela Web e propor uma abordagem ao Design da Interação que reduza tais barreiras. Barrei-ras semióticas são barreiBarrei-ras relacionadas à comunicação, mediação e representação. Estas barreiras surgem na colaboração mediada pela Web pois muitos mecanismos da comuni-cação interpessoal face-a-face não estão disponíveis. Dependendo do contexto, barreiras semióticas frequentemente exercem um impacto negativo à colaboração; entretanto, em alguns casos o impacto pode ser positivo também.

A abordagem ao Design da Interação aqui proposta tem suas bases na Web Pragmática e utiliza a Semiótica Organizacional e a Teoria da Atividade como referenciais teórico-metodológicos. As investigações teóricas contaram com uma contrapartida em termos de um embasamento em práticas reais através da participação em um projeto de pesquisa no domínio da educação inclusiva. Materializamos a abordagem proposta no design de um protótipo e na implementação de uma ferramenta correspondente ao protótipo, que apoia uma prática de profissionais no domínio da educação inclusiva. Além disso, propusemos e conduzimos um método de avaliação guiada pela pragmática dentro do contexto de um estudo de caso longitudinal. O design do protótipo, a implementação da ferramenta e a avaliação conduzida fornecem evidências de que a abordagem proposta ao Design da Interação guiada pela pragmática contribui para a redução de barreiras semióticas e para a promoção da colaboração mediada pela Web.

(12)

(13)

Acknowledgments

Luckily, writing a PhD thesis is not a solitary enterprise. During the last years, I had the privilege to collaborate or exchange ideas with many people who knowingly or unkowingly left their marks on this work. First and foremost, I would like to thank my orientadora Cecilia Baranauskas who inspired, challenged, and guided me, finding always the right dose and setting an example on a professional and personal level.

My thanks go to the dwellers of sala 71 and sala 90, “old” and “new”, who shared many a nice conversation and not less lunch, cups of tea or coffee: Alessandro, Elaine, Leonardo, Roberto, and Vanessa. I would like to thank the people who gave me the opportunity to participate as co-author of their work: Ana Carla, Carla, Diego, Elaine, Julio, Lara, Leonardo, Leonelo, Roberto, Rodrigo, and Vagner. There is hopefully a lot more to come. And of course, I’d like to thank all the others of InterHAD, an incredible group of people, for all the happy-hours, desmanches and other activities.

I’m grateful for the oportunity to participate in the “TNR” project, and to learn with and from colleagues and friends who challenged my thinking with different perspectives. Of course, this especially includes the semeadoras who participated in the evaluation re-ported in Chapter 6: Ângela, Andréa, Eliene, Lidinalva, Rosimeire, Simone, and Vanessa. I would like to thank CAPES, CNPq, Microsoft Research-FAPESP, NIED, the Insti-tute of Computing, and UNICAMP for financial support.

Last, but certainly not least, I would like to thank my parents and my brother for their support and understanding, and Katrin, the love of my life without whom nothing of this would have been possible, for her inconditional support and patience.

(14)

(15)

“All that I want is keeping it easy. It’s what I want that’s easy, It’s getting it that’s complicated.”

Torbjørn Brundtland, Svein Berge (Röyksopp)

(16)

(17)

6.5.4 On PDE . . . 114 6.6 Conclusion . . . 116 7 Conclusions 119 7.1 Contributions . . . 120 7.2 Critical Reflection . . . 121 7.3 Future Work . . . 123 Bibliography 127 A Conceptual Frameworks for Interaction Design: AT and OS 141 B Identifying Pragmatic Patterns of Collaborative Problem Solving 153 C Implementing the Timeline Representation in TNR 163 C.1 Main prototype characteristics . . . 163

C.2 Differences between prototype and implementation . . . 164

C.3 Implementation details . . . 166

C.4 Using the timeline . . . 168

C.5 Topics for evaluation and further design . . . 171 D Approval Certificate of the Ethical Review Board 175

E Permissions from Publishers 179

(22)

(23)

List of Tables

4.1 Juxtaposing the Syntactic, the Semantic and the Pragmatic Web. . . 45 5.1 Examples of timelines for facilitating different kinds of collaboration. . . . 68

(24)

(25)

List of Figures

1.1 Method as conducted in this thesis. . . 8 1.2 Distribution of the research method over the chapters of the thesis. . . 10 3.1 People access content independent of services but sensitive to context. . . . 36 4.1 The Web as a layered Information System. . . 47 4.2 The Pragmatic Web perspective on Interaction Design. . . 48 4.3 Interaction in the Pragmatic Web. Adapted from [71]. . . 49 4.4 Instantiated research and design process. . . 51 5.1 Wireframe of the timeline within the CMS. . . 74 5.2 Timeline canvas. . . 75 5.3 Chat widget. . . 76 6.1 “Our Cases” landing page (texts in Portuguese). . . 97 6.2 Overview of a case (texts in Portuguese). . . 98 6.3 Temporal sequence of the evaluation. . . 98 6.4 Controlled AEE case discussion. . . 99 6.5 Comments by role and case. . . 100 6.6 Flow between responses to comments by role. . . 101 6.7 Temporal flow of all comments. . . 102 6.8 Accesses of participants to case 1. . . 102 6.9 Comment flow during the activity. . . 104 6.10 A snapshot of the case discussion. . . 105 C.1 The timeline prototype. . . 164 C.2 The initial view of a case discussion’s timeline representation. . . 169 C.3 Hovering with the mouse pointer above a comment. . . 170 C.4 Comment details in context. . . 171 C.5 Accesses to the case discussion and author filter. . . 172

(26)

(27)

Chapter 1 Introduction

The Web has long since become a medium of interaction among people. People use Web-based services and applications in many different kinds of situations: for professional or personal purposes, at work, at home or en route, while working, learning, playing, relaxing, etc. While interacting they might encounter different kinds of barriers. Some of these barriers are topics of interest in the Human-Computer Interaction (HCI) field and related disciplines. These barriers refer to different aspects of the social relations among people, the applications they use, and the purposes and situations of use. For example, accessibility and usability both focus on barriers considering the relation between a person and the user interface of an application.

In this work, we are concerned with barriers of collaboration among people in the context of practices mediated by web-based systems. We use the word “collaboration” here to denote “working together with others toward a common goal”. With “practices” we mean the actual executions of purposeful actions. A practice is not only defined by its actions or purpose, but also includes the participants with their needs and preferences, the tools they use, as well as the situational context and cultural values. Practices are not necessarily defined formally or explicitly. People participating in a practice might have individual understandings of a practice. In order to successfully conduct a practice, there usually have to exist some common understandings of the practice. Furthermore, practices may change or evolve over time, as participants aggregate experience. In many contexts, practice evolution is explicitly desired, e.g. in educational systems, but also in business systems where efficiency or performance increases are important goals.

“Mediation” means that practices are not carried out directly (“immediately”) on the object of collaboration but mediated, for example using a representation of an object that needs to be interpreted. Thus, the medium has a strong influence on practice conduc-tion and might facilitate some acconduc-tions while making others more difficult. Changes in the medium might result in changes in practices and vice versa. We focus on web-based

(28)

2 Chapter 1. Introduction

systems as mediators of collaborative practices, i.e. we are especially interested in how tools (the functionalities provided by a web system) and the web-based system as an en-vironment mediate practices. The Web as a technological platform for mediation implies that collaboration can occur distributed in time and space, using different devices in dif-ferent situations. An arbitrary number of difdif-ferent people may collaborate, with difdif-ferent preferences and different degrees of knowledge about the practice as well as about each other.

As an example of a collaborative, web-mediated practice consider the collaborative writing of a scientific article. A primary object of collaboration is the article itself, in which for example research results are reported with the goal of disseminating or sharing knowledge. Secondary objects of collaboration might be the processes of submitting and publishing the article. Some parts of this practice are formally defined. For example, depending on where the article is to be published, there might be defined an upper limit of the article size, formatting guidelines or a submission deadline. Other parts of the practice might be implicitly defined. For example, an article reporting research results should at least explain the research problem, the objectives and the method, and present and discuss the results. Still other parts might not be defined at all, but might be negotiated before starting the practice (e.g. whether to use the camera-ready formatting template), or emerge during practice (e.g. the layout of figures and tables, used abbreviations). The authors might or might not have previous experience writing scientific articles individually or collaboratively, and might or might not know some of the co-authors. All this has an influence on how each author will contribute to the writing process, and depending on previous experiences, an author might approach collaborative writing differently than last time. Regarding mediation, different tools and environments favor different kinds of practice conduction. Round- or token-based editing on the one hand and real-time online collaborative editing on the other hand facilitate certain kinds of writing while complicating others.

This short example already illustrates some important aspects of web-based collab-oration relevant to this work and hinted at possible barriers. In the remainder of this introduction, we will present the problem we are treating in this work, our objectives and method, and present the outline of this thesis.

1.1 Problem

Web-mediated collaboration has three important aspects that are part of the problem treated in this work: communication, representation and mediation.

Generally, successful communication is a prerequisite of successful collaboration. An important element of communication among people is the interpretation of intentions of

(29)

1.1. Problem 3

the communication partners. During face-to-face communication, non-verbal clues such as facial expressions, gestures, inflection, etc. facilitate the interpretation of intentions. These clues and their interpretation are determined by linguistic, social, cultural, and other aspects which delimit the communication partners’ behavior and provide common ground. During written or computer-mediated communication these mechanisms are not always directly available, even when using video communication tools. Thus, commu-nication and consequently collaboration might be impaired and breakdowns might occur that need to be resolved. This situation even aggravates when considering that computer-mediated collaboration is often temporally and spatially distributed and that people often collaborate even not knowing each other very well.

During collaboration in the physical world, we can either directly access the object of collaboration or access a representation of it. When building a house or repairing a car we can directly access the construction site or the car, or we can work with floor plans, scale models, wiring diagrams, etc. Both direct access and representations support different actions differently, e.g. planning, troubleshooting, or fixing. On the other hand, during collaboration in the digital world, we always access representations. Note that even in the case of “direct manipulation” [114], we ultimately access representations of ideas, messages, stories, etc. Collaborating in the physical world, we are able to create some representations autonomously, using our own tools. Those who cannot create their tools usually have a choice of different tools, which might be customized. This is also true to some limited extent for collaboration in the digital world, where users have some choice, e.g. regarding hardware form factor, operating system, or software. However, in the Web, today most people depend on the tools others made, e.g. one might be able to use the web browser or search engine of ones choice, but in order to participate in a discussion forum, one usually still has to use the forum software the forum owner provides. These digital “tools” sometimes have restrictions regarding the task intended by the tool user, and oftentimes are actually representations themselves that need to be interpreted or “understood”.

Web-based collaboration is always mediated, i.e. any action such as manipulation of the object of collaboration or communication in order to coordinate manipulation is shaped by mechanisms or systems before it becomes effective and can be perceived by collaboration partners. Representations, tools, and the web-based systems themselves are examples of mediators. A mediator often shapes or transforms the actions or contents that it mediates, enabling or facilitating certain actions while making others more difficult or even impossible. A diagram as a visualization of a data set might lead to insights that are different from interpreting a table or the raw data. The opposite direction is also possible, i.e. the mediated also has an influence on the mediator. For understanding “big data”, i.e. very large and complex data sets, new visualizations are required.

(30)

The web-based system or environment in which collaboration takes place can be seen as a medium itself. Implicit and explicit norms valid in the system are mediators of the relationship between people, and members of the system can have an influence on the relationships between other members.

Another important aspect of mediation is that of flexibility, e.g. the possibility to choose or customize a tool and to use it the way one “sees best fit”. When engaging in collaborative practices, people have different needs and preferences with regard to which tools to use and how to manipulate or access shared artifacts. Musicians usually bring their own instruments to a jam session, construction workers might prefer one brand of tools over another. Novices might perform reasonably well using the instruments or tools they are accustomed to, but their performance might decrease dramatically when e.g. playing the guitar of a friend. In the cases of e.g. machine operators or aircraft pilots, even considerable training is required in order to be able to use another tool of the same kind. Experts are often able to switch tools more easily, but even then they might not be able to achieve the same level of performance. A player of a pipe organ usually needs some time to get accustomed to a new organ, because keyboards and pedalboard have different characteristics, stops have different positions and trigger different types of pipes, etc. Tool choice not only affects performance, but the overall user experience, and, most importantly for the context of this work, interpretation or “making sense of things”. Data presented in a table may facilitate different interpretations than the same data presented as a graph. For different kinds of data one representation might be more adequate than the other, and different people might prefer different representations.

In the case of web-mediated collaboration, due to limited interoperability and relative closedness of web systems, flexibility is very limited. When choosing a web application or system for collaboration, all participants have to use the same set of tools and represen-tations.

In summary, there exist barriers to web-mediated collaboration that are related to lim-itations regarding communication, representation, and mediation. These barriers make it more difficult to express and interpret intentions, i.e. to individually and collabora-tively construct meanings and successfully conduct practices and develop them further. It should be noted that the three types of barriers are not arbitrary choices. At the begin-ning of this study, we had no classification for the barriers described in this section. They were “just” barriers, some reported in scientific literature, some observed during practical experiences of participating in different research and design projects. After defining our theoretical and methodological frame of reference (Chapters 3 and 4) and after identify-ing Organizational Semiotics and Activity Theory as fundamental base in the research (Chapter 4 and Appendix A), we realized that these barriers were related to communi-cation, representation, and mediation, which are core concepts and units of analysis of

(31)

1.2. Objective 5

Organizational Semiotics and Activity Theory.

1.2 Objective

The barriers described in the previous section are essentially semiotic ones, i.e. they are related to how and why people produce, interpret and use signs during web-based collaboration. Thus, the main objective of this thesis can be formulated as follows:

“Investigate semiotic barriers to web-mediated collaboration, and propose an approach to Interaction Design that reduces these barriers.”

We have already presented our understanding of “web-mediated collaboration” and explained the barriers and other relevant aspects at the beginning of this introduction. The term “approach” in our objective is used in contrast to “method” and “technique”, borrowing from Anthony [6] in the context of English language teaching, and from Dix [43] in the context of theory creation in HCI: “techniques carry out a method which is consistent with an approach. [. . . ] An approach is axiomatic. It describes the nature of the subject matter to be taught. It states a point of view, a philosophy, an article of faith - something which one believes but cannot necessarily prove. It is often unarguable except in terms of the effectiveness of the methods which grow out of it” [6, p. 63f.]. “A theoretical approach is also not so much a method or technique that is applied to research, but an attitude and a desire to understand, in some ordered way, the phenomena around us. This approach can influence design and research methodology [. . . ]” [43, p. 175].

Of course, regarding a PhD thesis, the parts “[. . . ] something which one believes but cannot necessarily prove [. . . ]”, and “[. . . ] often unarguable except in terms of the effectiveness of the methods which grow out of it [. . . ]” are problematic. Thus, a secondary objective of this work must be:

“Demonstrate that it is possible to define an Interaction Design method that can be carried out using respective techniques and that results in a decrease of aforementioned barriers.”

In terms of scientific work, this is an “existence proof”, i.e. a proof that our proposed approach can in fact be instantiated by defining and carrying out a method, leading to meaningful results. A stronger, alternative objective would be to define a method and validate it, or “prove” that this method is at least as “good” as other related methods, where “good” would have to be defined in terms of some measurable criteria. Our PhD thesis is a qualitative as opposed to a quantitative work. In HCI, qualitative research and design methods can be validated by justification and evaluation [43]. Justification may use

(32)

expert opinion, published results, and argumentation based thereupon. Evaluation may use peer review, comparison with previous results, or empirical evaluation methods [47]. Empirical evaluation is meant as only one component that contributes to the evaluation part of validation and comprises different evaluations in different contexts by different evaluators. At UNICAMP, there is a four to six years time frame for concluding a PhD thesis. Our proposal of an approach to Interaction Design reached sufficient maturity at a later stage of our doctoral research. It would be quite unlikely to find different contexts and external evaluators that produced meaningful results given the typical four to six years time frame of a PhD project. Furthermore, basing the evaluation on a single empiric evaluation by ourselves would be methodologically quite unsound [43]. Hence, while this PhD thesis provides the justification part of validation, it cannot provide the evaluation part. Thus, we defined our secondary objective as an existence proof and not as a validation. In this context, please also note that the chapter about evaluation in this thesis is not a validation or evaluation of our approach, but an evaluation employing our proposed approach.

A pertinent question is, why our objective is to propose a new approach and not a new method within an existing approach? In recent years, different authors have noted that new challenges for HCI have arisen (e.g. [9, 29, 64, 115]). These challenges are related to the fact that the use of information technology is permeating all aspects of life and that people with an ever increasing diversity of characteristics interact with each other in an ever increasing diversity of situations. The challenges that emerge from our problem of web-mediated collaboration described in section 1.1 are well aligned with the challenges described by Bannon, Bødker, Harrison, and Shneiderman. In order to solve these chal-lenges, some authors claim that radical changes in HCI research and practice are required, e.g. Bødker identifies a third wave, Harrison a third paradigm of HCI. Bannon and Shnei-derman identify traces of novel HCI research and practice, and Bannon concludes with a call-to-action, “[. . . ] encouraging an openness to new forms of thinking about the human-technology relationship [. . . ]” [9, p. 57]. On this background, our decision for this PhD thesis was to propose and explore an approach to Interaction Design that is adequate for discussing contemporary challenges for HCI theory and practice.

1.3 Method

The main contribution of this PhD thesis lies in the development of an approach to Interaction Design research and practice as opposed to the development of methods and techniques within a theoretical framework or approach. Before presenting the method employed we thus must make explicit the ontological and epistemological assumptions that are the basis of this work. We subscribe to a neo-humanist paradigm [69], i.e. the

(33)

1.3. Method 7

view that the world is subject to change and conflict, and that knowledge about the world is subjective as opposed to objective. A subscription to the neo-humanist paradigm also entails a statement about our role as a researcher and designer: “The neohumanist paradigm seeks radical change, emancipation, and potentiality, and stresses the role that different social and organizational forces play in understanding change. It focuses on all forms of barriers to emancipation – in particular, ideology (distorted communication), power, and psychological compulsions and social constraints – and seeks ways to overcome them” [69]. This quote might seem radical to some readers. However, with the adoption of Participatory Design by HCI at the latest, it has become clear that research and design are not independent of social or political questions and that researching and designing also means to be aware of these questions and to take a position. We take the position described in this paragraph, a position that we believe is compatible with what Baranauskas [10] describes as “Socially Aware Computing”, and a position that we believe is aligned to Interaction Design as described by Löwgren and Stolterman [85].

The general method adopted for this PhD thesis consisted of the following elements: • characterization of the problem and objectives,

• literature review for theoretical grounding,

• subscription to a world view (epistemological and ontological basis),

• participation in research projects for practical grounding (empirical frame of refer-ence),

• definition of the theoretical and methodological frame of reference, • definition of our approach to solve the problem,

• application of the approach,

• analysis and partial validation, and • conclusion.

As already indicated by the last paragraph of section 1.1, this method was not exe-cuted in a sequential, waterfall-like order. Based on a continuous literature review and the epistemological and ontological basis (“world view”) of neo-humanism [69] we chose and continually refined a theoretical and methodological frame of reference. Further-more, an empirical frame of reference was taken for practical grounding. Our proposal of an approach to Pragmatics-driven Interaction Design was based on these frames of reference and was applied to a system design and evaluation in order to provide validity

(34)

Figure 1.1: Method as conducted in this thesis.

to the proposal. Figure 1.1 provides a graphical representation of how the method was conducted.

This PhD thesis is an articles-based thesis, i.e. with exception of this introduction and the conclusion, the chapters are scientific articles that have been published or submitted to publication. Furthermore, although the method elements follow a rough order, they do not form a rigid sequence of steps. Consequently, the sequence of elements of the adopted method does not match the sequence of chapters in this thesis. We have already described the problem and objectives of this work. Additional aspects of the problem and objectives can be found in the early chapters of this work. A critical literature review was an ongoing activity of this work and thus is part of most chapters of this thesis. The definition of the theoretical and methodological frame of reference for treating the problem has to be aligned to the epistemological and ontological assumptions. Further-more, the frame of reference has a crucial impact on the definition, application, analysis, and partial validation of our proposed approach. Important aspects of the problem and objectives of this work are related to linguistics and semiotics, and more specifically to pragmatics. Examples of these aspects are questions such as how and why people use language and other signs during collaborative meaning construction and negotiation in the Web. We thus chose the Pragmatic Web [111] as our theoretical and

(35)

methodologi-1.4. Outline of the Thesis 9

cal frame of reference. The Pragmatic Web is based on Organizational Semiotics (OS; [83]) and the Language/Action Perspective (LAP; [58, 131]). The Pragmatic Web per se makes no statement about Interaction Design. Thus, we had to complement our frame of reference with our view of Interaction Design that is compatible with Baranauskas’ [10] Socially Aware Computing.

The same way that the work was informed by and articulated with scientific literature, all research activities were grounded in practical experiences and data gathered during our participation in research and design projects. Two projects inspired and informed our work, and provided the means to apply and validate our theoretic findings. The first project was the e-Cidadania project [12], a project with the objective of studying and proposing inclusive solutions for exercising citizenship. This project served as motivation for this PhD thesis, since we realized that besides the barriers related to universal and participatory access [120] addressed by the e-Cidadania project, there are yet other bar-riers, namely the semiotic barriers addressed by this PhD thesis. The project also served as a background scenario for Chapter 2 of this thesis and for [75]. The second project was the project “Projeto Redes Sociais e Autonomia Profissional”1 _{(English: “Social Networks} and Professional Autonomy”). This project investigates how to facilitate lifelong learning of teachers in the field of special education in Brazil’s public school system. Within the project, the social network “Todos Nós em Rede”2 (TNR; English: “All of Us Networked”) is a system where special education teachers can socialize, share experiences and discuss matters related to work practices. At the time of writing this thesis, the project is ongoing and the TNR system is under development. The project served as a background scenario for Chapters 3 to 6 and was crucial for applying and partly validating our approach. For the sake of brevity, we sometimes use “TNR” in this PhD thesis as a homonym, denoting either the research project or the social network system.

Figure 1.2 relates the elements of the method with the chapters of this thesis. Smaller squares signify that the respective chapter only makes a minor contribution to the re-spective method element. Since the literature review and the grounding in practical experiences occurred along all activities of this work, we omitted these two elements from Figure 1.2.

1.4 Outline of the Thesis

As already stated, this thesis is organized as an articles set. With exception of this introduction and Chapter 7, all chapters have been published or submitted to publication. Chapters 2, 3, and 5 have been presented at scientific conferences and published in the

1_{http://www.nied.unicamp.br/tnr}

(36)

Figure 1.2: Distribution of the research method over the chapters of the thesis.

respective proceedings. Chapters 4 and 6 have been submitted to journals.

Chapter 2 puts the problem treated in this work into the context of Bødker’s third wave of HCI and investigates how the Pragmatic Web might contribute to Interaction Design. It focuses on two main aspects of the problem: meaning negotiation among people and its support by the action repertoire provided by web-based systems. Regarding meaning negotiation, the chapter presents related literature from the area of the Pragmatic Web. Regarding action repertoires, literature about affordances and actability [35] is presented and put into the context of Interaction Design in the Pragmatic Web. Using actual and possible examples from the e-Cidadania3 project, issues related to meaning representation and negotiation as well as action repertoires are discussed. These issues are structured using Stamper’s “organizational onion” [122], a conceptualization of an organization as an information system consisting of 3 layers: the informal, the formal, and the technical.

The main contributions of Chapter 2 are:

• an articulation of Bødker’s third-wave HCI with the Pragmatic Web,

• the conclusion that a further development of Interaction Design related con-cepts and theories is required to treat the presented issues, and

(37)

1.4. Outline of the Thesis 11

• the insight that the Pragmatic Web might contribute to the identification and definition of these concepts and theories.

Chapter 2 has been published as:

Hornung, H. and Baranauskas, M. C. C. (2009). An Interaction Design Perspective on the Pragmatic Web: Preliminary Thoughts. In I-SEMANTICS’09: Proceedings

of the 5th International Conference on Semantic Systems, 2-4 September 2009, Graz,

Austria, pp. 695–705.

Chapter 3 investigates how web-mediated interaction can be facilitated employing a Pragmatic Web perspective. It identifies barriers people encounter while interact-ing in the Web. These barriers are related to information relevance, information presentation, and flexibility of use. The chapter presents a conceptual model of interaction under a Pragmatic Web perspective. The model has been created by an-alyzing a scenario inspired by the TNR project using concepts from post-cognitivist HCI frameworks [78] and neo-humanist Information Systems frameworks such as OS and LAP. Its main building blocks are people who interact with each other accessing content via services flexibly and context-dependently. This vision is substantially different from a vision of web-based interaction characterized by user accounts, in-terface features, and document access that depend on a specific web application. The main contributions of Chapter 3 are:

• a first definition of our vision of “Interaction Design in the Pragmatic Web”, • the methodological and theoretical grounding of the research in post-cognitivist

HCI frameworks, neo-humanist IS frameworks, and Web Science [15], and • a conceptual model for understanding web-mediated interaction given the

pro-posed perspective.

Chapter 3 has been published as:

Hornung, H. and Baranauskas, M. C. C. (2011). Towards a Conceptual Framework for Interaction Design for the Pragmatic Web. In HCII’11: Proceedings of the 14th

International Conference on Human-Computer Interaction, Part I, 9-14 July 2011,

Orlando, FL, USA, pp. 72–81.

Chapter 4 presents a Pragmatic Web-based approach to conceptualizing and designing web-based applications for mediating social interaction. This approach is articulated with contemporary challenges for HCI theory and practice discussed in [9] and [115]. It builds upon the model presented in Chapter 3, a further analysis of related

(38)

conceptual frameworks (Organizational Semiotics and Activity Theory; cf. [73] and Appendix A), and an analysis of empirical data from the TNR project that identified pragmatic patterns of interaction ([74]; cf. Appendix B). The chapter instantiates the work presented in Chapter 3 using the TNR project as a concrete research and design project, and the cyclic Design and Research process proposed in [63]. It is shown that, by focusing on people’s actual practices and by explicitly considering the pragmatic dimensions of context, our vision of “Interaction Design in the Pragmatic Web” concretely contributes to the design of web-based collaborative systems. The main contributions of Chapter 4 are:

• a characterization of the Pragmatic Web,

• a more detailed, instantiated definition of our vision of “Interaction Design in the Pragmatic Web”, and

• evidence that the Pragmatic Web is a workable approach to the design of Web applications that facilitate collaborative meaning construction and negotiation. Chapter 4 has been submitted to a journal:

Hornung, H. and Baranauskas, M. C. C. (submitted). Pragmatics-driven Design of Web-Mediated Interaction. Submitted to a journal.

Chapter 5 applies the approach presented in Chapter 4 to create a low-fidelity prototype of a tool for web-based collaborative problem solving. The tool uses a timeline metaphor for information visualization and is supposed to facilitate the concrete practice of a “case discussion” within the context of the TNR project. Regarding problems in the scope of Interaction Design in the Pragmatic Web, the following aspects are addressed: flexible information presentation, facilitation of meaning negotiation, as well as flexible and creative practice conduction. The process of constructing the prototype covered the stages “understand”, “study”, and “design”. Subsequently, and not reported in Chapter 5, the prototype has been transformed into a functional prototype (cf. Appendix C). Thus the “build” stage is also covered. The main contributions of Chapter 5 are:

• application of “Pragmatics-driven design” in order to design a prototype, and • discussion of an existing information visualization metaphor under the

per-spective of “Interaction Design in the Pragmatic Web”. Chapter 5 has been published as:

(39)

Hornung, H. and Baranauskas, M. C. C. (2012). Timelines as Mediators of Lifelong Learning Processes. In IHC’12: Proceedings of the 11th Brazilian Symposium on

Human Factors in Computing Systems, 5-9 November 2012, Cuiabá, Brazil, pp.

99–108.

Chapter 6 proposes an approach to evaluation of web-based collaborative systems within Pragmatics-driven Design as described in Chapter 4. The approach is positioned as a goal-free evaluation within a continuous design-in-use cycle in which design and evaluation are understood as two sides of the same coin. The main characteristics of the proposed approach are: immersion of the designer-evaluator in the system to be evaluated, involvement of relevant stakeholders in the evaluation process, observa-tion of a real practice “in situ”, and a group activity for collaborative sense-making. The approach is instantiated within the context of the TNR project. The evalua-tion uncovered semiotic barriers and positive and negative tensions between these barriers. Furthermore, it informed design and redesign as well as future evaluations. The main contributions of Chapter 6 are:

• a description of Pragmatics-Driven Evaluation (PDE) and an application of the approach,

• a demonstration that PDE leads to a deeper understanding of phenomena occurring in the evaluated system,

• evidence that PDE can result in concrete implications for design, and

• additional validation of the pragmatics-driven approach proposed in this thesis by showing that PDE produces meaningful results.

Chapter 6 has been submitted to a journal:

Hornung, H. and Baranauskas, M. C. C. (submitted). Pragmatics-driven Evaluation of Web-Mediated Interaction. Submitted to a journal.

Chapter 7 concludes and presents a critical reflection as well as future work.

Appendices A and B as described in subsection 1.4.1 present some of the conceptual topics and their application in more detail.

Appendix C describes the high-fidelity prototype that corresponds to the low-fidelity prototype presented in Chapter 5.

Appendix D contains a copy of the Approval Certificate of the Ethical Review Board regarding the TNR project that serves as recurring example and case study to this thesis.

(40)

Appendix E contains the copies of the permissions from the respective publishers to include the papers that have been published elsewhere into this thesis.

During the course of our PhD project we published other work. The work related to this thesis is presented briefly in subsection 1.4.2.

1.4.1 Supplemental Texts

We included two chapters into the Appendix that discuss some topics in more detail and thus provide complementary information especially to Chapters 4 and 6. The first investigates conceptual frameworks that are compatible with the Pragmatics-driven design and evaluation presented in this thesis. The second applies the concepts in order to support the identification of recurring problematical or positive situations during web-mediated problem solving, a special case of web-web-mediated collaboration.

Appendix A investigates how to choose a conceptual framework as the theoretical frame of reference for a concrete Interaction Design problem. To guide this choice, filter criteria are proposed that are based on the socio-technical context of the design problem and the skills, attitudes and experiences of the involved people (previous projects, multidisciplinary mix, stance towards ontological, epistemological and ped-agogic questions). These filter criteria are applied to the context of the TNR project that also serves as recurring example and case study in this thesis. As a result Ac-tivity Theory (AT) and Organizational Semiotics (OS) are identified. The chapter presents and discusses both frameworks in the context of the TNR project and with respect to their support to different stages of Interaction Design. Appendix A has been published as:

Hornung, H. and Baranauskas, M. C. C. (2013). Conceptual Frameworks for In-teraction Design: Analysing Activity Theory and Organizational Semiotics Contri-butions. In ICISO’13: Proceedings of the 14th International Conference on

Infor-matics and Semiotics in Organisations, 25-27 March 2013, Stockholm, Sweden, pp.

136–146.

Appendix B presents a practical application of the theoretical concepts related to Prag-matics that are presented in this thesis. It is based on previous work that examined data gathered during so-called “scenarios”, participatory practices that examined different web-based systems with regard to their adequacy as systems to support practices in the context of the TNR project [24, 23]. In Appendix B we conduct a micro- and macro-pragmatic analysis to identify pragmatic patterns of collaborative problem solving. These patterns describe recurring situations of use which might

(41)

require design of solutions that facilitate, promote, or avoid the manifestation of the pattern. Appendix B has been published as:

Hornung, H., Bonacin, R., dos Reis, J. C., Pereira, R., and Baranauskas, M. C. C. (2012). Identifying Pragmatic Patterns of Collaborative Problem Solving. In

ICWI’12: Proceedings of the IADIS International Conference WWW/Internet,

18-21 October 2012, Madrid, Spain, pp. 379–397.

1.4.2 Related Work

The following three papers present work related to this thesis. The first two present micro-pragmatic analyses of data gathered during participatory practices conducted during the TNR project. They were informed by some theoretical concepts presented in this thesis and in turn provide some empirical basis to Chapters 4, 5, and 6. The third paper illustrates some of the concepts presented in Chapters 3 and 4 using the example of the inclusive social network “Vila na Rede” built during the e-Cidadania project.

Bonacin et al. 2013 perform an analysis of data collected during participatory prac-tices of the TNR project. The main object of this analysis is that of “dynamic knowledge” during collaborative problem solving, i.e. the meaning making and ne-gotiation processes. The analysis includes a pragmatic function analysis which is based on Semiotics and Speech Act Theory. The authors identify research chal-lenges and possible new interaction mechanisms that are enabled by knowledge about pragmatic aspects of interaction. They furthermore identify a preliminary research framework for the computational, conceptual and interactive dimensions of this problem. This paper is an extended version of [24] and has been published as:

Bonacin, R., Hornung, H., dos Reis, J. C., Pereira, R., and Baranauskas, M. C. C. (2013). Pragmatic Aspects of Collaborative Problem Solving: Towards a Frame-work for Conceptualizing Dynamic Knowledge. In Enterprise Information Systems,

volume 141 of Springer Lecture Notes in Business Information Processing (LNBIP),

pp. 410–426.

Bonacin et al. 2013 build upon the previous paper and present a model expressed in the Web Ontology Language (OWL). The OWL model can be used to represent some pragmatic aspects of communication. By instantiating the model using the data gathered during two participatory practices of the TNR project, the authors illustrate some possible information retrieval scenarios and discuss some limitations of a pragmatic communication analysis using formal semantic web techniques. This paper is an extended version of [22] and has been published as:

(42)

Bonacin, R., dos Reis, J. C., Hornung, H., and Baranauskas, M. C. C. (2013). An ontological model for supporting intention-based information sharing on collaborative problem solving. International Journal of Collaborative Enterprise, 3(2/3):130–150.

Hornung et al. 2013 illustrate the vision of “Interaction Design in the Pragmatic Web” using the example of the inclusive social network system “Vila na Rede”. The paper presents a further elaboration of some ideas presented in Chapter 3 and preliminary ideas that have been further developed in Chapter 4. Among others, it discusses the relation between inclusive access to information and the Pragmatic Web, as well as contextualized and customized interaction using the conceptual model of Chapter 3 on the example of “Vila na Rede”. This paper has been published as:

Hornung, H., dos Reis, J. C., and Bonacin, R. (2013). Sistemas inclusivos sob a ótica da Web Pragmática (in Portuguese). In Baranauskas, M. C. C., Martins, M. C., and Valente, J. A., editors, Codesign de redes digitais: tecnologia e educação

a serviço da inclusão social, chapter 14, pp. 275–293. Penso, Porto Alegre, RS,

(43)

Chapter 2 An interaction design perspective on

the Pragmatic Web: preliminary

thoughts

1

2.1 Introduction

During the last years, digital artefacts are being used in more and more diverse config-urations. Within the context of this paper a “digital artefact” is anything created by humans and accessed via computerized technology (e.g. a word processing application running on a local PC, an electronic government service accessed via a mobile device, digital interactive television). Not only professionals are interacting with digital artefacts in a purposeful manner, but people are accessing services anywhere, anytime for different purposes, be it in a work context or simply for entertainment or other leisure related activities. Bødker has coined the term “the third wave of HCI (Human-Computer Inter-action)” to characterize this broadening and intermixing of use contexts and application types [29].

This third wave can be best described via juxtaposition with the second wave: the third wave is about non-work contexts, non-purposeful or non-rational actions, etc. It focuses on the cultural level and expands the view from mere cognitive to emotional aspects. This shift to the third wave in the HCI discipline – although leaving open still unsolved issues of the second wave – poses many new challenges and questions regarding

1_{Article presented at I-SEMANTICS 2009 and published as “Hornung,} _{H. and}

Baranauskas, M. C. C. (2009). An interaction design perspective on the Pragmatic Web: preliminary thoughts. In Paschke, A., Weigand, H., Behrendt, W., Tochtermann, K., and Pellegrini, T., editors, Proceedings of the 5th International Conference on Semantic

Sys-tems (I-SEMANTICS ’09), Sept. 2–4 2009, Graz, Austria, page 695–705.”

(44)

18 Chapter 2. An interaction design perspective on the Pragmatic Web

interaction design.

We must not forget that a great number of potential users has difficulties to access those applications or services or has no access to them at all. Reasons for this are manifold: illiteracy, no experience in using digital artefacts, or special needs not attended by the service in question, only to name a few. In the context of Brazil for example, the Brazilian Computer Society has addressed this issue when defining the fourth of five “Grand Chal-lenges in Computer Science Research in Brazil” as “Participative and Universal Access to Knowledge for the Brazilian Citizen” [120].

The challenge refers to technological, educational, cultural, social and economical barriers to the access and the interaction with digital artefacts, whereas “access” is not only defined in the narrower sense of accessibility but in the more comprehensive sense of legibility, that addresses the problem of how to deliver information that makes sense and is relevant to users.

With regard to applications or services accessed via the Internet, today’s HTML-based so-called Syntactic Web does not offer many mechanisms to facilitate the under-standing of content apart from content formatting and structuring. For example, the Wikipedia article on “Semantic Web” contains a thumbnail of the W3C’s Semantic Web logo (http://en.wikipedia.org/wiki/Semantic_web), but apart from the logo caption,

nothing indicates that the depicted image is the logo of a W3C activity related to the subject of the Wikipedia article. Furthermore, it is not clear, how the article and the activity are related (e.g. if the article is a summary of the activities key findings, if the activity is an example of an institution incidentally working on the same subject as the article’s authors, etc.)

The Semantic Web has been proposed as an extension to the current Web with the intent to introduce meaning to Web pages, processable by human or machine agents [16]. Currently, many languages exist (e.g. RDF Schema; http://www.w3.org/TR/rdf-schema), or the Web Ontology Language (OWL; http://www.w3.org/TR/owl-features/)), that al-low for knowledge modelling and meaning sharing, and that provide a basis for semantic interoperability [113]. However, the augmented semantic contents to a great extent remain inaccessible or unintelligible for human agents.

Gandon illustrates some of the problems of retrieval of semantically annotated in-formation and proposes a mechanism to facilitating the interpretation of query results [52]. However, he does not treat questions related to interaction design. Despite this and other efforts having been made to make the Semantic Web intelligible for humans, according to McCool it will never achieve widespread adoption “because it’s a complex format and requires users to sacrifice expressivity and pay enormous costs in translation and maintenance” [87, p. 86]. However, Singh claims that the vision of the Semantic Web can be implemented via Pragmatics, a branch of Semiotics that deals with context-based

(45)

2.2. Related Work – Pragmatic Research and the Pragmatic Web 19

meaning [118]. The purpose of this paper is thus to investigate how the vision of the Pragmatic Web can contribute to the interaction design of services that are accessible, intelligible and relevant, whereas – in contrast to the Semantic Web – the interaction happens primarily between human agents. The paper is organized as follows: section 2.2 gives an overview of relevant literature in adjacent areas, section 2.3 proposes an initial approach to conduct interaction design in the Pragmatic Web, and discusses the unveiled issues, section 2.4 concludes.

2.2 Related Work – Pragmatic Research and the

Pragmatic Web

The challenges that arise with the third wave of HCI require a socio-technical view. Orga-nizational Semiotics (OS) is a discipline that recognizes the need to approach the question of how to facilitate and promote access to digital artefacts in a way that considers tech-nical as well as social issues [124]. It focuses on understanding the different properties of signs on various levels to analyse and design information systems in terms of three hu-man information functions: expressing meanings, communicating intentions and creating knowledge.

The primary focus of this paper lies in the pragmatic aspects of digital artefacts, i.e. intentions, motivations, negotiations or, in other words, the different actions that are possible to be executed on a digital artefact and the question why and how actions are executed. Pragmatics deals with intentions, communications, conversations, negotiations, etc., i.e. with the purposeful use of signs. Important concepts with regard to Pragmat-ics are the “pragmatic information”, i.e. the personal knowledge and experience of each communicating partner, the shared knowledge (that is higher if the partners are from the same cultural community), and the context where the communication takes place, whereas the context is comprised of elements such as speaker, hearer, intention, purpose, theme, time, location, etc. [83, chapter 3.2].

Besides Organizational Semiotics there exist other approaches that are suitable to investigate these questions of users acting in a broader organizational context, whereas organization is defined in a broader sense than for example the work context. Cordeiro and Filipe [32] compare the Language/Action Perspective (LAP), Organizational Semiotics (OS) and the Theory of Organized Activity (TOA) and propose an integration of the three approaches into a combined one.

Cordeiro and Filipe [32] and Goldkuhl [55] show that there still exist many open ques-tions with regard to conceptualization, explanation and understanding with the help of different action oriented or pragmatic theories. Theories adapted from reference

(46)

disci-20 Chapter 2. An interaction design perspective on the Pragmatic Web

plines have different strengths and weaknesses. The Language/Action Perspective for example presumes that actions are purposeful, neglects tacit communication and has a focus on a professional work context [130].

The integration of LAP, OS and TOA proposed by Cordeiro and Filipe starts from a TOA activity and is based on a vocabulary mapping and on drawing analogies between the respective key concepts [32]. Within their integration model, the human within his social domain is acknowledged as the central concept. OS’es information fields like family, religion or country expand the activity domain.

LAP, TOA and OS are often applied to business contexts that substantially differ from Bødker’s third-wave HCI. In this context users that interact with each other might be influenced by entirely different information fields, thus in the worst case, the only implicit shared context that can be assumed is the URL accessed via the browser. Actions are not always rational; often the primary purpose of an action is entertainment or distraction.

Regarding the analysis of actions and action repertoires, depending on the theoretical frame of reference, different concepts exist. Gibson defined, “the affordances of the envi-ronment are what it offers the animal, what it provides or furnishes, either for good or for ill” [54, p. 177]. An affordance is “something that refers to both the environment and the animal”. Affordances in the sense of Gibson are inherent properties that simply exist. Norman [95] introduced the term into design in 1988, referring himself to “perceivable affordances”, i.e. properties that do not simply exist, but that have to be perceived. Since then, the term has been widely adopted by the HCI community, not always in the sense originally intended by Norman. In a later essay, he introduces the term “(social) signi-fier” that should replace the term “affordance”. The term “social signisigni-fier” reflects the social character of most actions that we perform and on the other hand allows accidental signifiers: “the perceivable part of an affordance is a signifier and if deliberately placed by a designer, it is a social signifier” [95, p. 19].

In OS, affordances are used in the original sense of Gibson: they are invariant reper-toires of behaviour and as such constitute the perceivable reality of a human agent [124]. Thus, Norman’s “social signifiers” can be seen as an approximation to the affordances as defined in OS.

“Actability” is a similar concept that is concerned with social actions mediated by information systems. It has been defined as “an information system’s ability to perform actions, and to permit, promote and facilitate the performance of actions by users, both through the system and based on information from the system, in some business con-text” [35, p. 1076]. Goldkuhl provides a comparison of “actability” with “affordances” in Gibson’s sense [56].

From the perspective of interaction design, social signifiers, affordances and actabil-ity are concepts that are concerned with a pragmatically oriented action repertoire, i.e.

(47)

2.2. Related Work – Pragmatic Research and the Pragmatic Web 21

the different actions a human agent can perform on a digital artefact. “Social signifiers” and “actability” seem to have a closer relation to the human-computer interface, whereas “affordances” in the OS sense emphasize the social context of the human agent. Further-more, “affordances” exist independently of the digital artefact and thus seem to be an appropriate concept to be already considered in the phase of artefact creation.

Regarding the implementation of the Pragmatic Web vision using the theoretical con-structs described above, examples from literature often refer to web services. A main difference of the approaches described below and our interaction design perspective is that web services often refer to electronic and not to human agents. However, many results below can be mapped or adapted to our case.

With relation to web services, Singh identifies challenges that cannot be addressed by merely considering the Syntactic or Semantic Web, but that require a pragmatic view of the problem [118]. For example, web services cannot be fully described by the methods they provide; instead, a model that permits the negotiation between service provider and consumer about if and how to interact with each other would be more adequate. Singh lists three principles of pragmatic web service design: user before provider, process before data, and interaction before representation [118].

Although Singh is concerned with web services provided and consumed by electronic agents, the challenges and principles he identified can be partly matched to human agents. E.g., the principle “interaction before representation” that refers to hiding “excess” se-mantics when describing the interaction specifications of services clearly applies to the modelling and implementation of ()action repertoires in the human-computer inter-face [118].

Another milestone in the relatively recent history of the Pragmatic Web is the paper by Schoop et al., in which the Pragmatic Web vision is defined as “to augment human collaboration effectively by appropriate technologies, such as systems for ontology ne-gotiations, for ontology-based business interactions, and for pragmatic ontology-building efforts in communities of practice” [111, p. 76]. Scientific contributions in the area reveal different understandings and accentuations of the term “Pragmatic Web”.

De Moor and van den Heuvel adopt a semiotic perspective to examine how virtual communities can pragmatically select web services [40]. Since the purpose of their work is theory construction, it has to be examined whether and how this methodology can be adopted to our usage scenario. On the other hand, it has to be questioned, if their approach can be adapted to be applicable by communities of “non-expert”/non-technical users. Furthermore, it is not clear, how the challenges identified by Singh [118] are addressed.

Liu explores the context of pragmatic web services [84]. He describes methods of how to construct, discover and rank pragmatic web services in order to be able to use the

(48)

22 Chapter 2. An interaction design perspective on the Pragmatic Web

right service at the right time. Although the author works with computational agents, some of the concepts might be mapped to human agents. For example, a pragmatic web service can be interpreted as an element of the action repertoire. For a given user, alternative candidates may exist that all yield the desired result, but that may be more or less appropriate.

De Moor addresses the problem of modelling context in the Pragmatic Web and pro-poses pragmatic patterns for the meaning negotiation processes. Meanings are assigned to syntactic resources and formalized in ontologies [37]. Meaning alignment is concerned with the compatibility of ontologies. Ontologies are modelled on the semantic and used on the pragmatic layer: agents in the Pragmatic Web select meanings and meaning rep-resentations and negotiate meanings among each other.

In our approach, we adopt the conceptual model of the Web by de Moor [37] with semantic resources in the Semantic Web layer and a set of pragmatic contexts in the Pragmatic Web layer. The semantic resources exist mainly in the form of ontologies, i.e. we agree with de Moor and use a mix of large, detailed, standard ontologies and independent, domain-specific micro-ontologies. These are supplemented by potentially large folksologies, i.e. ontologies that contain an unrestricted and non-stable set of entities and that are generated by amateur users in an uncoordinated way [121].

2.3 A view on interaction design in the Pragmatic

Web

The previous section has shown that the Pragmatic Web is a still emerging area with different ideas and approaches, many of which build on the Syntactic Web and are con-cerned with (semi-)automated agents or expert human agents. The goal of this section is to explore how the concepts of the Pragmatic Web can be applied to the interaction of non-expert human agents and what design questions arise in this context. “Non-expert human agents” include users with special needs, low or no literacy skills and no or low computer skills.

Regarding non-expert human agents, the following questions arise: How can users se-lect meanings from already existing ontologies? How can users create or adapt meanings? What are good meaning representations? How can users negotiate meanings with other users or non-human agents? These questions are interrelated and have a common denom-inator: How can users benefit from more legible and more relevant information without having to worry about the conceptual aspects of the Pragmatic Web. For example, we cannot expect users to construct meanings querying and aligning different existing on-tologies. End users should not be concerned with reading ontology diagrams or other

Interaction design in the pragmatic Web = reducing semiotic barriers to web-mediated communication

Heiko Horst Hornung

“Interaction Design in the Pragmatic Web –

Reducing Semiotic Barriers to Web-mediated

Collaboration.”

“Design da Interação na Web Pragmática –

Reduzindo Barreiras Semióticas na Colaboração

Mediada pela Web.”

CAMPINAS

2013

Interaction Design in the Pragmatic Web – Reducing

Semiotic Barriers to Web-mediated Collaboration.

Heiko Horst Hornung

Abstract

Resumo

Acknowledgments

Contents

List of Tables

List of Figures

Chapter 1

Introduction

1.1

Problem

1.2

Objective

1.3

Method

1.4

Outline of the Thesis

1.4.1

Supplemental Texts

1.4.2

Related Work

Chapter 2

An interaction design perspective on

the Pragmatic Web: preliminary

thoughts

1

2.1

Introduction

2.2

Related Work – Pragmatic Research and the

Pragmatic Web

2.3

A view on interaction design in the Pragmatic

Web